research Archives - Amateur Earthling

How safe is our bike infrastructure?

A major study in Vancouver, BC has looked at the relative safety of biking on various types of infrastructure, by looking at several hundred actual accidents resulting in hospital visits by cyclists over an 18 month period. Vancouver has a lot of different types of bike infrastructure, some of it extremely safe, so it’s a good test bed. Unsurprisingly, busy roads with parked cars and no on-street bike facilities were the most dangerous. Cycletracks (physically separated dedicated on-street bikeways) were an astonishing 90% safer. Interestingly, multi-use paths weren’t all that much safer, despite being preferred by many cyclists.

Sequencing Risks

A reporter from Bloomberg joins the PGP, only to discover that he carries a rare and potentially pathogenic acquired mutation, found in people with blood disorders. He doesn’t know how to deal with it, and neither do the doctors, really. Which is half of what this study is all about. What do we need to learn as a society to deal with knowing our sequences? A lot, I’m guessing.

More Roads = More Traffic

A new study from the University of Toronto clearly shows that additional free road capacity — either from adding actual road, or shifting people from driving to transit — has no effect on congestion. Traffic expands to fill the available capacity, no matter how much you add, and the net public benefit from the investment in additional road capacity is negative.

PLoS ONE: The Network of Global Corporate Control

The Network of Global Corporate Control, as revealed by a research group at ETH Zurich (kind of the Swiss MIT). Their core finding: a densely connected “super entity” of 147 corporations, about 75% of which are financial intermediaries, has an amount of control which is ~10 times as great as their economic scale would suggest. They are all majority owned by by other members of the super entity, and so have co-incident interests, and are likely to act as a bloc to protect themselves and each other, and to be subject to simultaneous group failure when under duress. Not like this is a huge surprise or anything, but it’s nice to see it described in a quantitative framework.

The capitalist network that runs the world

A team at the Swiss equivalent of MIT has revealed a dense knot of power and ownership interconnections within a particular subset of the world’s transnational corporations. It will come as no surprise that these companies are overwhelming financial firms… but this is the first time that anyone has really been able to lay out the structure of this network of power that runs the world in detail, accounting for all of the subsidiary ownerships and mutual shareholding. Tyler Durden would be inspired. The research paper will be published in PLoS One here Real Soon Now.

Solving protein structures with social computation

In just 16 days gamers solved a protein structure that had eluded computational methods for a decade. Distributed social computing? What other computationally difficult problems are susceptible to this kind of approach?

Discounted cashflow analysis of scientific programming

Software Carpentry does a little math describing the value of teaching scientists how to build good software. Even with very pessimistic assumptions, it’s clearly worthwhile. With realistic assumptions, it’s a frigging research bonanza. WTF? Why don’t advisers and administrators make sure everyone is on board?

Tag: research